/*
Mirta Galesic/Edward Cokely	
Proposal title: Influence of Affect on Philosophical Intuitions


HYPOTHESES

Stated-Hyp1: "We hypothesize that people have more positive moral
evaluations of actions done "with virtue" as opposed to actions that maximize good
consequences and follow the correct moral rule" (p. 2)

	// Condition removed from fielded survey. Cannot test.

Stated-Hyp2: "we hypothesize that virtuous actions will be evaluated more positively independently of whether the behaviors in question are helpful (characterized by positive affect) or harmful (characterized by
negative affect)." (p. 3)

	// Condition removed from fielded survey. Cannot test. The original question is on p. 8-9 of the proposal (appendix)

Stated-Hyp3: "We hypothesize that people should be more likely to have responses consistent with moral objectivism in a scenario involving high affect compared to a scenario involving low affect"

	// Low affect condition was removed from fielded survey. Cannot test the hypothesis. See page 9 of the proposal.
	
Stated-Hyp4: "We hypothesize that people should have higher ratings for freedom and moral responsibility in a high affect scenario compared to a low affect scenario"

	// Low affect condition was removed from fielded survey. Cannot test the hypothesis. See page 9 of proposal.
	
Stated-Hyp5: "we hypothesize that the behaviors with harmful side effects
(characterized by negative affect) will be judged as more intentional than behaviors with helpful side effects (characterized by positive affect)."

	Test-Hyp1: Perceptions of intentionality will be higher in the case of behaviors with harmful side effects vs. helpful side effects.
	

********************************************************************************
NOTES:

- For test hypothesis 6, we conducted a paired ttest because all respondents 
answered both questions.

- Low affect condition was removed from the moral objectivism case. Cannot test 
stated hypothesis 3.

- In the free will case (stated hyp 4), the tax condition was removed from the 
survey so we cannot test this hypothesis. 
*/

clear all
use "Galesic090.dta", clear

********************************************************************************

* INDICATORS OF EXPERIMENTAL MANIPULATIONS


* OUTCOME MEASURES

* moral objectivism
	replace K8=. if K8==-1
	clonevar objectivism=K8


* intentionality
	replace K5=. if K5==-1
	replace K6=. if K6==-1
	
	clonevar intent_harm =K5
	clonevar intent_help =K6
	
	tab intent_harm, mis 
	tab intent_help, mis
 	

********************************************************************************

* ANALYSIS
	
	

*Test-Hyp1: Perceptions of intentionality will be higher in the case of behaviors with harmful side effects vs. helpful side effects.
	
	ttest intent_harm == intent_help
	tess +, ttest init(Galesic90) 
	// do not reject. 0.0000
